sign in sign up
<<
Home , Joey Votto, Zack Cozart

Supplemental Materials

Survey Questions

Please choose your favorite team.   Colorado Rockies  Los Angeles Dodgers   St. Louis Cardinals   Baltimore Orioles  Boston Red Sox   Cleveland Indians  Detroit Tigers  Houston Astros  Kansas City Royals  of Anaheim   New York Yankees  Oakland Athletics  Seattle Mariners  Tampa Bay Rays  Texas Rangers  Blue Jays Instructions

In this study we want to know what you think about sports teams. You will see 8 MLB games that are being played tonight, grouped into 2 sets of 4 games each. For each game we will remind you of a few facts about each of the two teams and then ask you a few questions about the game. Here are the questions. Which team would you like to see win? How much would you say that you care about who wins this game? Answer this question on a scale from 0 to 100, where 0 = Not at all (I couldn’t care less. I wouldn’t even bother to check which team won.) and 100 = the most I could care about any game ever (If my team loses I’ll be devastated, and if they win I’d be totally thrilled. I care so much that whether my team wins or loses, I’ll never forget the result.) Please predict the winner to the best of your ability and knowledge. This is the team that you believe will win, regardless of whether you want them to win or whether you believe that they should win. Express your confidence as a probability from 50 to 100 where, 50 = 50-50 chance, complete toss-up to 100 = Absolutely certain that the team you predicted would win WILL win. Please begin with the first game.

Example of Presentation of games

Chicago Cubs (43-61, 23-26 home; Jackson, 5-11, 5.76 ERA) versus Colorado Rockies (43-62, 16-34 away; De La Rosa, 11-6, 4.19 ERA) (Presented if favorite team was in the )

Chicago White Sox (51-55, 24-31 away; Quintana, 5-7, 3.15 ERA) versus Detroit Tigers (57-45, 26-25 home; Sanchez, 7-4, 3.45 ERA) (Presented if favorite team was in the American League)

Preference Question

Which team would you like to see win?  Chicago Cubs  Colorado Rockies  Chicago White Sox  Detroit Tigers

Strength of Preference Question How much would you say that you care about who wins this game? Answer this question on a scale from 0 to 100, where 0 = Not at all (I couldn’t care less. I wouldn’t even bother to check which team won.) 100 = the most I could care about any game ever (If my team loses I’ll be devastated, and if they win I’d be totally thrilled. I care so much that whether my team wins or loses, I’ll never forget the result.) (Please use the slider to indicate your preference. Even if the slider is automatically where you want it, you must click it with the mouse for your choice to register.) ______

Prediction Question (Preference and Prediction questions were on separate pages – game information was present for both questions)

Chicago Cubs (43-61, 23-26 home; Jackson, 5-11, 5.76 ERA) versus Colorado Rockies (43-62, 16-34 away; De La Rosa, 11-6, 4.19 ERA) (Presented if favorite team was in the National League)

Chicago White Sox (51-55, 24-31 away; Quintana, 5-7, 3.15 ERA) versus Detroit Tigers (57-45, 26-25 home; Sanchez, 7-4, 3.45 ERA) (Presented if favorite team was in the American League)

Please predict the winner to the best of your ability and knowledge. This is the team that you believe will win, regardless of whether you want them to win or whether you believe that they should win.  Chicago Cubs  Colorado Rockies  Chicago White Sox  Detroit Tigers

Express your confidence as a probability from 50 to 100 where, 50 = 50-50 chance, complete toss-up 100 = Absolutely certain that the team you predicted would win WILL win. ______Likelihood that the (Preferred team was entered here) will win.

Consistency Prime Instructions CRITICAL REASONING On the next page you will get a challenge: to explain a conflicting set of facts, and to do so in only 3 minutes. Please work for the full 3 minutes. Provide explanations that go beyond the "obvious" answer. If you really can’t come up with even one good answer, then tell us by typing something like "I just couldn't explain why." Please answer thoughtfully. Every answer will be read. It is important that you find thoughtful, genuinely sensible explanations. The page will automatically change after exactly 3 minutes - it is okay if it changes while you are typing.

Consistency Prime Conundrum Why do most people today strongly reject prejudiced social beliefs from a hundred years ago on intrinsic grounds, even though there is basically no intrinsic difference between people today and people from the beginning of the last century? (Please type your answer into the text box below - the page will automatically change after 3 minutes. It is fine if you do not complete your response.)

Control Instructions CRITICAL REASONING On the next page you will get a double challenge: to explain a set of facts, and to do so in only 3 minutes. Please work for the full 3 minutes. Provide explanations that go beyond the "obvious" answer. If you really can’t come up with even one good answer, then tell us by typing something like "I just couldn't explain why." Please answer thoughtfully. Every answer will be read. It is important that you find thoughtful, genuinely sensible explanations. The page will automatically change after exactly 3 minutes - it is okay if it changes while you are typing.

Control Question Why do most people today strongly reject prejudiced social beliefs from a hundred years ago? (Please type your answer into the text box below - the page will automatically change after 3 minutes. It is fine if you do not complete your response.)

Delay (Mind Clearing Task) Instructions On the next page you will see a passage about the evolution of the horse. There is a minimum of three minutes for you to spend reading the passage. Read carefully as there will be a reading comprehension check at the end of the survey. Delay Task Evolution of the Horse: The evolution of the horse occurred over a period of 50 million years, transforming the small, dog-sized, forest-dwelling Eohippus into the modern horse. Paleozoologists have been able to piece together a more complete outline of the modern horse's evolutionary lineage than that of any other animal. The horse belongs to the order Perissodactyla (odd-toed ungulates), the members of which all share hooved feet and an odd number of toes on each foot, as well as mobile upper lips and a similar tooth structure. This means that horses share a common ancestry with tapirs and rhinoceroses. The perissodactyls arose in the late Paleocene, less than 10 million years after the Cretaceous–Paleogene extinction event. This group of animals appears to have been originally specialized for life in tropical forests, but whereas tapirs and, to some extent, rhinoceroses, retained their jungle specializations, modern horses are adapted to life on drier land, in the much-harsher climatic conditions of the steppes. Other species of Equus are adapted to a variety of intermediate conditions. The early ancestors of the modern horse walked on several spread-out toes, an accommodation to life spent walking on the soft, moist grounds of primeval forests. As grass species began to appear and flourish, the equids' diets shifted from foliage to grasses, leading to larger and more durable teeth. At the same time, as the steppes began to appear, the horse's predecessors needed to be capable of greater speeds to outrun predators. This was attained through the lengthening of limbs and the lifting of some toes from the ground in such a way that the weight of the body was gradually placed on one of the longest toes, the third. In June 2013, a group of researchers announced that they had sequenced the DNA of a 560–780 thousand year old horse, using material extracted from a leg bone found buried in permafrost in Canada's Yukon territory. Prior to this publication, the oldest nuclear genome that had been successfully sequenced was dated at 110–130 thousand years ago. For comparison, the researchers also sequenced the genomes of a 43,000 year old Pleistocene horse, a Przewalski's horse, five modern horse breeds, and a donkey. Analysis of differences between these genomes indicated that the last common ancestor of modern horses, donkeys, and zebras existed 4 to 4.5 million years ago. The results also indicated that Przewalski’s horse diverged from other modern types of horse about 43,000 years ago, and had never in its evolutionary history been domesticated. Subjective Knowledge: Please rate how well each statement describes you on the scale below. 1 - 2 - 3 - 4 - 5 - 6 - 7 - Strongly Disagre Moderatel Neither Moderatel Agre Strongl Disagre e y Disagree Agree y Agree e y Agree e Nor Disagre e I am interested in major        league baseball I have good knowledg e about        My knowledg e about major league        baseball is higher than my peers

Baseball Knowledge Test

This player is the active leader in career on-base percentage (minimum of 3000 plate appearances).   Albert Pojols   Alex Rodriguez What is the distance between the front of the pitcher's rubber and the rear point of home plate?  54 ft. 0 in.  56 ft. 6 in.  58 ft. 0 in.  60 ft. 6 in.

Which active pitcher has given up the most home runs?  CC Sabathia  Bronson Arroyo  Mark Buehrle  Bartolo Colon

This former Baltimore Orioles star holds the record for most consecutive games played at 2,632.  Cal Ripken Jr.   Lou Gehrig 

The last realignment of MLB divisions was before the 2013 season, seeing this team switch leagues.  Houston Astros  Seattle Mariners  Colorado Rockies  Philadelphia Phillies

This team has won 2 of the last 5 World Series Championships, led by manager Bruce Bochy.  St. Louis Cardinals  San Francisco Giants  Boston Red Sox  New York Yankees

This Cincinnati Reds player set a minor league single season record for steals with 155 in 2012.   Vince Coleman  Billy Hamilton For the 2013 season, this player had the lowest earned run average among qualified pitchers.   Felix Hernandez 

According to Baseball-References Wins above Replacement metric, this current player produced more value through his age-21 season than any player in history.  Miguel Cabrera   Alex Rodriguez 

The metric ERA+ adjusts a pitcher’s ERA (earned run average) according to the pitcher’s ballpark and the league’s run scoring environment. ERA+ is normalized so that a score of __ reflects the league average.  0  4  50  100

Attention Check Think back to the passage on the evolution of the horse. Analysis of differences between genomes indicated that the last common ancestor of modern horses was also the ancestor of two other equids. Name at least one of them.

Frustration measurement How frustrating did you find having to spend a minimum of three minutes looking at the conundrum? ______1 Not frustrating at all to 7 Extremely frustrating (continuous slider scale)

Suspicion Check

Was there anything in the tasks or information in the study that made you suspicious?  Yes  No

If yes, please indicate in the space below what you believe you noticed. Was there anything in this study that did not make sense, or did not seem to belong?  Yes  No

If yes, please indicate in the space below what you believe you noticed.

Demographics

You are:  Male  Female

Your age is (in years):

You are:  White  Black/ African American  Asian Indian  Chinese  Filipino  Japanese  Korean  Vietnamese  Other Asian (FILL IN): ______ Native American/ American Indian/ Alaskan Native (FILL IN Tribe): ______ Native Hawaiian or Other Pacific Islander  Mixed Ethnicity (example: Chicano and Native American, FILL IN): ______ Other (FILL IN): ______

Are you of Hispanic, Latino, or Spanish origin?  No, not of Hispanic, Latino or Spanish origin  Yes, Mexican, Mexican American, Chicano  Yes, Puerto Rican  Yes, Cuban  Yes, Central American (FILL IN): ______ Yes, South American (FILL IN): ______ Yes, Spanish (Spain) What is the highest level of education you have completed?  Less than High School  High School / GED  Some College  2-year College Degree  4-year College Degree  Masters Degree  Doctoral Degree  Professional Degree (JD, MD)

Is English your first language?  Yes  No

Are you a citizen of the United States?  Yes  No

How much did you enjoy participating in this experiment?  1 - Not at all  2  3  4  5  6  7 - Very Much Supplemental Material Additional Analyses Part 1 – Analyses including Non- Favorite Games

DB Analysis with Non-Favorite and Favorite Games We report an analysis including DB of both favorite and non-favorite games that also treats Prime and Order as separate factors (unlike the main analysis, which simplifies the comparisons by combining unprimed/control games from both response orders). All significance tests are two- tailed. The overall analysis is a 2 Prime X 2 Order X 2 Favorite mixed model controlling for Win Percentage and including participant and games as random factors. This analysis yielded a main effect of Favorite (F(1, 5091.03) = 138.67, p < .001), with a higher DB for favorite games (M = 14.34, SE = 0.94) compared to non-favorites/distractors (M = 4.43, SE = 0.59) and a main effect of Order (F(1, 2553.91) = 8.98, p = .003), with a larger DB for Preference-first games (M = 10.82, SE = 0.80) compared to Prediction-first games (M = 7.95, SE = 0.82). The analysis also yielded 2-way interactions between Favorite and Order (F(1, 5086.32) = 9.07, p = .003) and Order and Prime (F(1, 5648.64) = 12.19, p < .001). The difference between Preference-first and Prediction-first games was statistically significant for Favorite games

(MPreference First = 17.03, SE = 1.22; MPrediction First = 11.64, SE = 1.27; F(1,5511.78) = 10.84, p =

.001) but not for Non-Favorite/distractor games (MPreference First = 4.60, SE = 0.69; MPrediction

First = 4.26, SE = 0.69; F(1,1102.39) = .21, p = .64). The difference between Preference- first and Prediction-first games was statistically significant for primed games (MPreference First

= 11.76, SE = 1.19; MPrediction First = 5.83, SE = 1.23; F(4868.30) = 14.04, p < .001) but not for control games (MPreference First = 9.88, SE = 0.80; MPrediction First = 10.07, SE = 0.80; ; F(1, 2393.31) = .04, p = .84). As shown in Table S1, these 2-way interactions were moderated by a 3-way Favorite X Order X Prime interaction (F(1, 5387.56) = 10.03, p = . 002), in which the effect of Order was restricted only to the primed Favorite games (F(1,5633.79) = 15.44, p < .001)). Unprimed Favorite games showed no effect of Order (F(1,5470.44) = 0.06, p = .82). Furthermore, DB was significantly smaller in primed Prediction-first Favorite games compared to unprimed Prediction-first Favorite games (F(1,5511.37) = 11.57, p = .001), however the difference between Preference-first primed Favorite games and controls fell short of significance (F(1,5519.07) = 2.62, p = .12). No significant differences were found for non-favorite games for primed games. For simplicity of exposition, we also performed the same analysis without the Favorite games. That is we analyzed only the non-favorite/ distractor games for any effects of Prime and Order. This 2x2 analysis yielded neither significant main effect nor a significant interaction. The four means for the non-favorite/distractor games shown in TableS1 are also plotted in Figure 2. Table S1. DB Score Means and SE for Each Condition Condition Order Favorite Control Primed Preference First Favorite 15.23 (1.21) 18.84 (2.0) Non-Favorite 4.52 (0.69) 4.68 (.10) Prediction First Favorite 15.60 (1.21) 7.69 (2.1) Non-Favorite 4.54 (0.69) 3.97 (0.93) Adjusted DB Score Favorite and Non-Favorite Analysis We repeated the prior analysis for the DB score on our adjusted DB score. These results were consistent with the primary DB measure (that takes into account model predictions). There was a significant effect of Favorite, with adjusted DB higher for favorite games. There was a significant effect of Order, with Prediction-first games showing lower adjusted DBs, F(1,5098.00) = 95.79, p < .001. As before, there were significant Order X Favorite (F(1,5095.21) = 15.78, p < .001) and Order X Prime (F(1,5654.99) = 5.33, p = .021) interactions, and a significant 3-way Favorite X Prime X Order interaction, F(1,5395.02) = 7.58, p = .006. As shown in Table S2, primed Prediction-first Favorite games had a significantly lower adjusted DB as compared to unprimed Prediction-first Favorite games (F(1,5581.06) = 8.54, p = .003) and primed Preference-first Favorite games (F(1,5639.80) = 13.74, p < .001), but did not differ from non-favorite games. Table S2. Adjusted DB Score Means and SE for Each Condition Condition Order Favorite Control Primed Preference First Favorite .49 (.06) .59 (.10) Non-Favorite -.06 (.03) -.09 (.04) Prediction First Favorite .39 (.06) .03 (.11) Non-Favorite -.05 (.03) -.05 (.04)

Prediction Accuracy (Raw Score) Favorite and Non-Favorite Analysis Next, we examined the role of priming and order with regard to prediction accuracy with a Prime X Order X Favorite logistic mixed model predicting accuracy (i.e., whether the participant correctly predicted the outcome of the game). This model yielded a significant Prime X Order interaction (F(1, 5656) = 4.03, p = .045), with greater accuracy for Prediction-first primed predictions (M = 0.55, SE = 0.03) as compared to Prediction- first control predictions (M = 0.49, SE = 0.03; F(1, 5656) = 2.97, p = .09). Preference- first primed accuracy (M = 0.49, SE = 0.04) was not significantly different from control (M = 0.53, SE = 0.03; F(1, 5656) = 1.23, p = .27). The 3-way interaction missed significance (F(1, 5656) = 3.28, p = .07), however, as shown in table S3, the overall pattern conformed to prior results. Table S3. Prediction Means and SE for Each Condition Condition Order Favorite Control Primed Preference First Favorite .54 (.04) .48 (.06) Non-Favorite .51 (.02) .50 (.03) Prediction First Favorite .50 (.04) .61 (.06) Non-Favorite .49 (.02) .48 (.02)

Corrected Prediction Accuracy Favorite and Non-Favorite Analysis We also examined the role of priming and order with regard to prediction accuracy with a Prime X Order X Favorite linear mixed model predicting corrected accuracy (i.e., participants’ accuracy minus TeamRankings). This model yielded a significant Prime X Order interaction (F(1, 5649.03) = 11.96, p = .001), with greater accuracy for Prediction- first primed predictions (M = -0.04, SE = 0.09) as compared to Prediction-first control predictions (M = -0.16, SE = 0.09; F(1, 5649.04) = 8.48, p = .004) and Preference-first primed predictions (M = -0.16, SE = 0.09, F(1, 5649.04) = 6.13, p = .013). Preference- first primed predictions were marginally less accurate as compared to Preference-first control predictions (M = -0.09, SE = 0.09; F(1, 5649.04) = 3.85, p = .05), and Preference-first control predictions were significantly more accurate than Prediction-first controls (F(1, 5649.03) = 7.11, p = .008). The 3-way interaction was also significant (F(1, 5649.07) = 6.06, p = .014), with more accurate predictions for favorite Prediction- first primed games than the favorite Prediction-first control (F(1, 5649.04) = 7.73, p = . 005) and favorite Preference-first primed (F(1, 5649.05) = 5.83, p = .016) games. Furthermore, Preference-first control predictions were marginally more accurate than Preference-first primed games for favorite games (F(1, 5649.05) = 2.77, p = .096). Finally, non-favorite Prediction-first control games were significantly more accurate than favorite Prediction-first control games (F(1, 5649.58) = 3.93, p = .048). Table S4. Corrected Prediction Means and SE for Each Condition Condition Order Favorite Control Primed Preference First Favorite -.09 (.09) -.21 (.10) Non-Favorite -.10 (.09) -.12 (.09) Prediction First Favorite -.20 (.09) .01 (.11) Non-Favorite -.12 (.08) -.10 (.09) Brier Score Favorite and Non-Favorite Analysis For individual predictions (as modeled here), the Brier score is the squared difference between one’s probability estimate and the actual outcome (0 or 1), which rewards high- confidence correct predictions and penalizes high-confidence incorrect predictions. Lower Brier scores indicate better accuracy. We conducted a linear mixed model predicting Brier scores with a 2 Prime X 2 Order X 2 Favorite design, including participant and game as random factors. This model yielded a main effect of Prime, with larger scores for control games as compared to primed games (MControl = 0.29, SE = .01;

MConsistency = 0.27, SE = .01; F(1, 5649.92) = 5.28, p = .022). There was also a Favorite X Order X Prime interaction in which Brier scores for primed Prediction-first favorite games were significantly lower as compared to both favorite Prediction-first unprimed games (F(1,5650.08) = 7.36, p = .007), and as compared to non-favorite Prediction-first primed games (F(1,5652.69) = 5.91, p = .015; see Table S5 for means and SEs). Table S5. Brier Score Means and SE for Each Condition Condition Order Favorite Control Primed Preference First Favorite .29 (.02) .28 (.03) Non-Favorite .29 (.01) .28 (.01) Prediction First Favorite .31 (.02) .23 (.03) Non-Favorite .29 (.01) .29 (.01)

DB Table with Condition and Strength of Preference with all games Table S6. Condition and Strength of Preference Predicting Desirability Bias for All Games Predictor B SE t p [95% CI] Intercept 5.90 0.44 13.42 < .001 [4.98, 6.81] Pref-First 0.69 .85 0.81 .42 [-0.99, 2.37] Pred-First -1.58 0.85 -1.86 .063 [-3.24, 0.08] Preference .10 0.01 10.23 < .001 Strength [.08, .12] Pref-First X .06 0.03 2.29 .022 Preference [.01, .11] Strength Pred-First X -.05 0.03 -1.81 .071 Preference [-.10, .003] Strength Note. Pref = Preference; Pred = Prediction. Both Preference and Prediction refer to primed games. The Strength of Preference variable was mean-centered (M = 44.967) for this analysis. Supplemental Material Additional Analyses Part 2 – Results for Analyses Separating Prime and Order Conditions for Favorite Games Only

DB Results for Favorite Games Only We conducted a 2x2 ANOVA Prime (Control, Consistency) X Order (Preference-first, Prediction-first), controlling for the win percentage. The analysis revealed a main effect of Order (F(1, 703) = 9.33, p = .002, partial eta-squared = .013), which indicated greater DB for Preference-first games (M = 11.46, SE = 1.28) as compared to Prediction-first games (M = 16.89, SE = 1.23). This effect was moderated by Prime (F(1, 703) = 10.95, p = .001, partial eta-squared = .013), such that DB for Preference-first games was only significantly greater than Prediction-first for consistency-primed games, F(1, 703) = 13.19, p < .001, partial eta-squared = .018 (see Figure 2). Furthermore, DB for Preference-First primed games was significantly greater than DB for Preference-first control games, F(1, 703) = 3.89, p = .049, partial eta-squared = .005, and DB for Prediction-first primed games was significantly smaller than DB for Prediction-first controls, F(1, 703) = 7.23, p = .007, partial eta-squared = .010. Table S7. DB Means and SE for Prime X Order Condition Order Control Primed Preference First 14.46 (1.21) 19.31 (2.14) Prediction First 14.92 (1.21) 7.99 (2.27) Note. Means and SE’s adjusted for Win-Percentage

Adjusted DB Results for Favorite Games Only A Prime X Order ANOVA (controlling for win-percentage) was performed on this adjusted DB score. These results mirrored the analysis with the prior DB measure, with greater Preference-first adjusted DB (M = 0.56, SE = .07) than Prediction-first (M = 0.23, SE = . 07), F(1, 703) = 11.38, p = .001, partial eta-squared = .016. As before, the effect of Order was moderated by Prime (F(1, 703) = 6.69, p = .01, partial eta-squared = .009), such that Preference-first DB was only larger in the primed condition (F(1, 703) = 11.12, p = .001, partial eta-squared = .016). Although the difference between Preference-first primed adjusted DB was greater than control games, this difference failed to reach significance in this analysis, F(1, 703) = 1.89, p = .17, partial eta-squared = .003. Prediction-first primed games’ adjusted DB was significantly lower than Prediction-first controls, F(1, 703) = 4.39, p = .037, partial eta-squared = .006. Table S8. Adjusted DB Means and SE for Prime X Order Condition Order Control Primed Preference First 0.47 (0.07) 0.66 (0.12) Prediction First 0.38 (0.07) 0.08 (0.13) Note. Means and SE’s adjusted for Win-Percentage Prediction Accuracy with Raw scores for Favorite Games Only A Prime X Order logistic regression predicting participants’ accuracy in predicting winners yielded a significant interaction between factors, χ2(1) = 3.88, p = .049. Participants in the Prediction-first primed condition was more accurate than those in the corresponding control condition as well as those in the Preference-first primed condition (contrasts for both comparisons missed significance with p = .07 and p = .08, respectively). Table S9. Prediction Means and SE for Prime X Order Condition Order Control Primed Preference First 0.54 (0.30) 0.48 (0.55) Prediction First 0.50 (0.30) 0.62 (0.54)

Prediction Accuracy Corrected by TeamRankings’ Baseline Accuracy for Favorite Games Only A Prime X Order ANOVA predicting participants’ corrected accuracy scores (participants’ accuracy minus TeamRankings’ accuracy for each condition) yielded a significant 2-way interaction, F(1, 704) = 10.82, p = .001, partial eta-squared = .015. Accuracy was greater in the Prediction-first primed condition (.06 above the proportion of correct predictions made by TeamRankings) compared both to the Prediction-first control condition (F(1, 704) = 8.54, p = .004, partial eta-squared = .012) and to the Preference- first primed condition (F(1, 704) = 6.76, p = .010, partial eta-squared = .010). The Preference-first primed prediction were marginally less accurate than corresponding Preference-first controls (F(1, 704) = 2.91, p = .089, partial eta-squared = .004). Unlike prior results, which indicated no difference between Orders in the control condition, Preference-first control predictions were significantly more accurate than Prediction-first controls (a reversal of the effect in the prime condition, F(1, 704) = 4.40, p = .036, partial eta-squared = .006). Table S10. Corrected Prediction Means and SE for Prime X Order Condition Order Control Primed Preference First -0.06 (0.04) -0.18 (0.07) Prediction First -0.17 (0.04) 0.06 (0.07) Brier Scores for Favorite Games Only Recall that lower Brier scores indicate greater accuracy. A Prime X Order ANOVA predicting Brier scores yielded a marginal main effect of Prime (F(1, 704) = 3.55, p = .06, partial eta-squared = .005), such that primed participants (M = .26, SE = .02) were more accurate than unprimed (M = .30, SE = .02). Despite the interaction failing to reaching statistical significance (F(1, 704) = 1.84, p = .18, partial eta-squared = .003), the main effect was largely driven by the Prediction-first primed condition (F(1, 704) = 5.04, p = . 025, partial eta-squared = .007). Table S11. Brier Score Means and SE for Prime X Order Condition Order Control Primed Preference First 0.30 (0.02) 0.28 (0.03) Prediction First 0.30 (0.02) 0.23 (0.03)